Summary

  • W3C issued DOM Level 1 on 1 October 1998 as a platform- and language-neutral API for accessing and changing HTML and XML documents.
  • The standard described logical document objects, their relationships and operations, while allowing products to keep different internal representations. Level 1 laid out Core and HTML interfaces; events remained for later levels.

A Web page could be displayed, but software needed a way to get its fingers around the parts. One browser's script might know a private route to a paragraph; an editor might carry a different map of the same document. The W3C DOM Level 1 Recommendation supplied public handles. A program could traverse a document, find an element, change its content and build new pieces through common interfaces for HTML and XML. The document stopped being a picture that each tool had to poke in its own dialect.

The Recommendation's introduction puts the challenge plainly. A document has a logical structure much like a tree. Its nodes have identities, relationships and functions. Two conforming implementations representing the same document should expose the same structural model. But neither must store its data internally as a tree. One cabinet might be built of wood and another of copper pipes; the drawers had to present the same handles to the person opening them.

That separation made the standard useful beyond browser tricks. The introduction traces the initial aim to portable scripts and programs across Web browsers, then names HTML/XML editors and document repositories as participants in the effort. DOM Core supplied common document interfaces, with HTML-specific interfaces built on top. The specification described them in language-neutral IDL and provided bindings for Java and ECMAScript. IDL was a way to state the contract precisely, not a demand that every product install the same internal object system.

The interfaces dealt in real operations. Programs could build documents, navigate a node's surroundings and add, modify or delete elements and content. A table cell could be reached as an object with a place in the structure rather than a patch of pixels on a screen. Such power matters to authors and readers because the same page can be inspected by a browser, edited by a tool and analyzed by another program. Common handles make the work portable at the level where document pieces actually live.

The 1 October 1998 W3C announcement called DOM Level 1 a standard API for dynamically manipulating documents and data. The word Level did honest work. This first installment covered document structure and content; the introduction reserved events, access control, thread safety, stylesheet rendering control and schema validation for later work. A script could get a hold of the page's furniture. The doorbell and the building's security desk had separate specifications to come.

The lasting historical move was a boundary drawn around the document as software sees it. HTML and XML remained their own languages. Products remained free to build different interiors. The DOM offered the same objects and behaviors at the exposed edge, so a tool could meet a document through a shared model rather than an undocumented favor from one application. In 1998, the Web page finally came with handles sturdy enough for more than one pair of hands.